An object becomes interrupted on a system if a SQL statement succeeds on the active system but fails with an error on the standby system. This mismatch causes objects to become interrupted on the system with the error. Business Continuity Manager replays the SQL statement to attempt recovery. Some errors may recover automatically, although others require manual intervention.
- A GRANT statement was run on the active system but not on the standby system, causing a permissions violation on the standby system.
- Standby system running out of perm space for a query.
- Standby system running out of spool space for a query.
- An external load job failed and left a load lock on a table on the standby system.
- An external cleanup script dropped an object on the standby system.
If an object is interrupted on a system, and another SQL statement or session needs to use that object, any other objects used by that session or SQL statement also become interrupted. This is a cascade interrupt. Once the objects or sessions required for this object recover, the cascade interrupts recover as well.
The recovery may take a few attempts, but if no more root cause interrupts occur, the objects become active.